Forced Turnover:
Evaluating Pressing Effectiveness in Soccer

Natalie Rayce

Carnegie Mellon University

David Almona

Centre College

Turning Defense to Attack

Manchester City wins back possession seconds after losing it through pressing.

Main Question

Can the effectiveness of a press in soccer be predicted using factors such as spatial context, pressing dynamics, game context and situational factors?

Key Terms

Pressing: a defensive tactic where players apply coordinated pressure on the opponent with the ball to force mistakes, win back possession, and quickly transition to attack

Forced Turnover: when a player loses possession due to opponent pressure, resulting in the opposing team gaining control. This includes misplaced passes, interceptions, successful tackles, or losing control under pressure - all direct results of effective defensive pressure

Data

  • Dataset: 520 matches in the MLS 2023 season

  • Three data types:

    • Match information: game details (teams, pitch, referee)
    • Event data: player actions (passes, shots, tackles) with timestamps and coordinates
    • Tracking data: real-time positions of all players and ball at 10 Hz
  • Source:

The Pressure Zone

  • Be within 6 meters of the ball carrier

Problem: 6-meter radius doesn’t account for direction

  • Equal threat from all angles?
  • Oversimplifies pressing

Circular Pressure Zones Ignore Directional Threats

Problem: 6-meter radius doesn’t account for direction

  • Equal threat from all angles?
  • Ignores player orientation
  • Oversimplifies pressure dynamics

The Pressure Zone: Revised

Methods: Detecting Pressure Sequences

  • Pressing actions were grouped into sequences if at least one defender continued pressing, allowing gaps of up to 1.5 seconds.
  • A new sequence was defined after any pause longer than 1.5 seconds.
  • For each sequence, the following features were recorded:
    • duration, number of defenders involved and average defender velocity at the start

Feature Engineering

  • Spatial Context: Ball carrier position, distance to boundaries, field third
  • Pressing Dynamics: Number of defenders, approach velocity, passing options
  • Game Context: Score, game state (winning/losing/drawing), time remaining
  • Situational Factors: How ball carrier gained possession (pass reception, interception, etc.), incoming pass characteristics (distance, height, range)

Modeling

  • Model:
    • Logistic Regression
    • XGBoost
    • All models evaluated using 10-fold cross-validation

Model Performance

plot

Results

plot

plot

Discussion

  • Limitations:
    • Inaccuracy in SkillCorner data
    • Individual player skill and tactical tendencies not considered in the model
    • Our model does not account for pitch control
  • Future Works:
    • Incorporate pressing intensity calculations
    • Incorporate pitch control